Friday, March 03, 2006
Generative Methodology Glass Bead Games
On the limits of the OWL standard à [184]
Reading material [1]
Reading material [2]
Reading material [3]
Summary of the discussion up to this point à [186]
On ontological modeling of biological expression
Subtitle: on the nature of T-Boxes and A-Boxes
Related note on OWL Full and
OWL DL reasoners
On Formal verses Natural systems à [206]
Comment from Irene
Paul,
Regarding your most recent post on n-ary relationships:
1. I am not able to understand what patterns are suggested by Dan on http://biopaxwiki.org/cgi-bin/moin.cgi/OntologyDesignPatterns
because these pattern discussions does not have the actual OWL implementation making them, in my view, impossible to assess.
2. As far as the W3C note on n-ary relationships, I do not think it concerns itself at all with 'class versus instance versus possibility of a resource being both, a class and an instance' issue.
You may be reading too much into a somewhat hard to parse sentence. [4]
Basically, all it is saying that sometimes we want to express something like Jack and Jill got married on September 5, 2004. (They are using a different example, of course). Everything in this example is an instance. In order to express it we need to introduce a new instance (individual) - Jack and Jill's Wedding. à
Paul’s comment: see the next two beads [213] and [214]
The above is data. The ontology that will support such expression will have classes. In addition to class Person to which, the individuals, Jack and Jill belong, we will now need a new class to which Jack and Jill's Wedding can belong. Let's say a class of Weddings.
So in this example 'got married' has been reified - it became a class.
We will need relationships as well. Such as, Wedding hasGroom Person, Wedding hasBride Person, Wedding hasDate xsd:date.
Returning to the different topic of class/instance distinctions, OWL Full, of course, allows a resource to be both an instance and a class.
DL algorithms have something they call Tbox (terminology about classes) and Abox (assertions about individuals), these are different algorithms. One deals with classes (Tbox) , another one with instances (Abox). They need for a set of classes to be disjoint with the set of individuals (instances).
Hope this help to clarify,
Irene
[1] http://dip.semanticweb.org/documents/ECIS2005-A-Methodology-for-Deriving-OWL-Ontologies-from-Products-and-Services-Categorization.pdf
[4] Comment by Paul Prueitt: I agree that I am reading to much in, but the question is about how one might do something…. And in this case the suggestion is that an “instance” needs to become a “class, and this is something that is not done in description logics, if I am reading this correctly.